The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were identified. The skill's operations are consistent with its documented purpose of providing an enhanced interface for the Nylas API.
[EXTERNAL_DOWNLOADS]: The skill installs the nylas-pp-cli tool via npx from the @mvanhorn/printing-press package. This resource originates from the verified namespace of the skill's author.
[COMMAND_EXECUTION]: Executes subcommands of nylas-pp-cli to sync, search, and manage email and calendar data. Destructive actions are protected by mandatory confirmation or a --dry-run capability for payload review.
[PROMPT_INJECTION]: The skill processes external content (emails and events), creating an indirect prompt injection surface. This risk is managed through several mechanisms:
Ingestion points: Data entering the context via sync, search, and since commands in SKILL.md.
Boundary markers: The --agent flag enforces machine-readable output formats, reducing unintended instruction following.
Capability inventory: Command execution via nylas-pp-cli and network delivery options.
Sanitization: Idempotency flags, dry-run previews, and read-only SQL modes provide layers of protection against malicious inputs.

pp-nylas